Under certain conditions, ruminal biohydrogenation produces trans fatty acids (FA), such as trans-10, cis-12 conjugated linoleic acid (CLA), and trans-10 18:1 (T10), which are associated with a phenomenon known as milk fat depression syndrome (MFD). The effect of milk trans-10 FA on MFD in small ruminants has been quantified using different regression models. However, no study has defined a regression that produces the most accurate statistical model. In addition, interspecies differences based on the relationship between milk trans-10 FA and milk fat content are yet to be evaluated. Therefore, we aimed to determine the best model that quantifies the relationship between MFD and trans-10 FA (CLA or T10) in goats and ewes, and to test the hypothesis that there is a species-specific difference between goats and ewes in the magnitude of milk fat loss by one unit change of milk CLA and/or T10 concentration. Previously published data (106 observations for goats and 68 for ewes) were used to estimate, by simple linear and nonlinear regressions, the following response variables: milk fat yield (FY; g/d), percentage change in milk fat yield (CFY;%), milk fat concentration (FC;%), and percentage change in milk fat concentration (CFC;%). The models with FC as response variable showed the best prediction adjustment for both species. However, the nonlinear relationship between FC and T10 proved to be the most accurate model, whereas CLA was the best regressor for ewes. In addition, there was no species-specific difference in the magnitude by which milk fat was reduced as a function of the milk CLA and T10 content.